Draft at 23 October 2013

Strategic review of the global WHO-coordinated invasive bacterial

vaccine preventable diseases and rotavirus sentinel hospital

surveillance networks, 2013

Note:

SAGE members have received two documents in the Yellow Book that summarize the

strategic review process, including the main findings and conclusions.

This draft document is provided should SAGE members require additional

information on the strategic review.

2

TABLE OF CONTENTS

Section Page

Preface 3

Acknowledgements 3

Abbreviations 4

Executive Summary 5

Informal Technical Advisory Group for New Vaccines Surveillance: Assessment

of the 2013 Strategic Review

13

1. Background - The Global IB-VPD and RV surveillance networks

- Strategic review of the IB-VPD and RV surveillance networks

18

2. IB-VPD surveillance network data analyses - Methods

- Summary of findings

3. RV surveillance network data analyses - Methods

- Summary of findings

4. IB-VPD and RV Laboratory networks review - Methods

- Summary of findings

- Conclusions and recommendations of the independent consultation

5. Data landscape analysis - Methods

- Summary of findings

- Conclusions and recommendations of the independent consultation

6. Use of surveillance network data - Literature review

- GAVI applications review

- Regional Office input

7. Ministry of Health perspectives - Methods

- Summary of finding

8. WHO roles and responsibilities

9. Surveillance networks funding

10. Summary and recommendations

References

Appendices

A. iTAG review team B. Data analysis plan for the strategic review

3

C. NUVI Strategic Review Meeting agenda

D. Data Landscape Analysis Report

E. IB-VPD and RV Laboratory Network Independent Review

4

PREFACE

Since 2008, the World Health Organization (WHO) has coordinated two global surveillance

networks for invasive bacterial-vaccine preventable diseases (IB-VPD) and rotavirus disease

(RV). The past five years have been dedicated to generating high-quality data to provide

countries and regions with quality data to help answer disease burden and vaccine impact

questions and assist with the planning of public health programmes. Given the finalization of the

2011-2020 Global Vaccine Action Plan and the continued acceleration of new vaccine

introductions, including pneumococcal conjugate and rotavirus vaccines, WHO and global

partners undertook a strategic review of the global IB-VPD and RV surveillance networks to

identify strengths and weaknesses and decide on strategies and actions to ensure that the

networks are responsive to the information needs of all immunization stakeholders.

The first section of this report of the strategic review provides background on the IB-VPD and RV

surveillance networks and the objectives and processes of the strategic surveillance networks

review. Subsequent sections describe the methods and main findings from the following eight

components of the review:

1. IB-VPD surveillance network data analyses

2. RV surveillance network data analyses

3. IB-VPD and RV laboratory networks review

4. Data landscape analysis

5. Use of surveillance data

6. Ministry of Health perspectives

7. WHO roles and responsibilities

8. Surveillance networks funding

The final section will summarize the conclusions and recommendations discussed by the Informal

Technical Advisory Group (iTAG) members and WHO Headquarters and Regional Offices.

ACKNOWLEDGEMENTS

The IB-VPD and RV networks depend on the dedication and efforts of staff at sentinel hospital

surveillance sites and laboratories, National Laboratories, Regional Reference Laboratories,

Global Reference Laboratories, and all levels of WHO. The strategic review was successfully

conducted due to these personnel who, in addition to normal duties, also provided information

and feedback as requested for the review. The members of the WHO informal Technical

Advisory Group for New Vaccines Surveillance generously gave their expertise ad time to help

guide the strategic review. WHOs global partners similarly devoted substantial time and efforts

to assist with the review process.

5

ABBREVIATIONS

ADIP Accelerated Development and Introduction Plans

AFR African Region of the World Health Organization

AMR Americas Region of the World Health Organization

AMS antimicrobial susceptibility

CRF case report form

ELISA enzyme-linked immunosorbent assay

EMR Eastern Mediterranean Region of the World Health Organization

EQA external quality assessment

EUR European Region of the World Health Organization

GRL Global Reference Laboratory

Hib Haemophilus influenzae type b

IB-VPD invasive bacterial-vaccine preventable diseases

iTAG informal Technical Advisory Group

MoH Ministry of Health

NL National Laboratory

NUVI New and Underutilized Vaccines Implementation

PCR polymerase chain reaction

PCV pneumococcal conjugate vaccine

QA quality assurance

QC quality control

RRL Regional Reference Laboratory

RV rotavirus

RO Regional Office

SEAR South-East Asian Region of the World Health Organization

SOP standard operating procedures

TWG Technical Working Group

VPD vaccine preventable disease

WHO World Health Organization

WHO HQ World Health Organization Headquarters

WPR Western Pacific Region of the World Health Organization

6

EXECUTIVE SUMMARY

Genesis of the strategic review Since 2008, WHO has been coordinating global sentinel hospital surveillance networks for invasive bacterial vaccine preventable diseases (IB-VPD) and rotavirus (RV) to provide quality data to assist with planning of public health programs around vaccine introduction and use. The objectives set for the network in 2008 were: During the pre-vaccine introduction period:

Document presence of disease, describe the disease epidemiology and provide data for estimating disease burden;

Establish a system to measure impact after vaccine introduction; Identify circulating serotypes and measure serotype distribution; and Monitor antibiotic sensitivity.

During the post-vaccine introduction period also: Assess disease trends over time; Monitor vaccination program impact; Monitor changes in circulating strains/serotypes; and Develop a platform for effectiveness and safety evaluation.

The longer term vision has been to establish sentinel surveillance for selected vaccine preventable disease (VPD) in low- and middle-income countries to complement existing active surveillance for polio, measles/rubella and passive surveillance and reporting for other VPDs. In February 2013, following 5 years of network coordination, WHO launched a full strategic review of both networks with its informal Technical Advisory Group for new vaccines surveillance (iTAG) to identify the networks strengths and weaknesses and to prioritize future actions. The iTAG assessed the networks capacities to document vaccine impact via the accelerating introductions of pneumococcal conjugate and rotavirus vaccines in the light of the call for quality case-based surveillance in the 2011-2020 Global Vaccine Action Plan. Recognizing the resource constraints in the targeted countries and competing health priorities, it was deemed imperative to determine best strategies to target human and financial resources. At the reviews outset, the iTAG and WHO recognized that while great strides had been made in the development of both IB-VPD and RV global surveillance networks, improvements were still required to ensure strong and effective networks that meet the ultimate goal of providing reliable data. IB-VPD surveillance was particularly recognized to be more complex than RV surveillance due to lack of adequate methods for laboratory confirmation of pneumococcal pneumonia, the rarity of meningitis as compared with diarrhoea, the difficulties of specimen collection and transport, and the complex bacterial laboratory diagnostic testing required. Prior to this review, in November 2012, WHO had established the following criteria for IB-VPD surveillance, in order to prioritize sites for support: Presence of a country surveillance management team, consisting of focal points at Ministry of

Health (MoH), sentinel hospital, sentinel hospital laboratory, and for data management; Countries conducting only Tier 1 meningitis surveillance enrol at least 100 suspect meningitis

cases[1] in children 38.5 C rectal or 38.0 C axillary) and one of the following signs: neck stiffness, altered consciousness with no other alternative diagnosis, or

other meningeal sign OR very patient aged under 5 years of age hospitalized with a clinical diagnosis of meningitis. [2]

Countries that do not enroll 100 suspect meningitis cases during 2013 but that significantly increased the number of enrolled cases over 2012 will be considered for funding on a case-by-case basis.

7

Objectives and methodology of the 2013 strategic review

The strategic review was conducted under the oversight of the iTAG that has been advising WHO on the conduct of sentinel surveillance since July 2011. The iTAG and WHO agreed on the following objectives for the strategic review: 1. Assess whether and to what extent the original 2008 objectives for the networks were met. 2. Critically assess the Ministry of Health (MoH) perspective, laboratory networks, and WHO

activities for capacity building, funding, and data management and the use of data; and 3. Provide conclusions and recommendations for strengthening the networks including defining

any needed complementary approaches for IB-VPD surveillance, and critically assess the networks future utility as a platform for other vaccine preventable disease (VPD) surveillance.

The strategic review consisted of several methods including analysis of surveillance data; questionnaires to obtain MoH perspectives; external review of laboratory function; external review of data management; review of the literature and GAVI applications to evaluate whether WHO surveillance data has been used in decision making; internal review of WHO activities and funding disbursement. Analysis of surveillance data was conducted on all data available to WHO, including IB-VPD case-based data collected by 4 Regional Offices (ROs) and data aggregated by sentinel site in 2 ROs. For RV, data for analysis was aggregated by sentinel site. Data from over 90 databases were consolidated and cleaned. IB-VPD variables were mapped across Regions, and the first uniform IB-VPD data analysis performed. Due to the volume of data and paucity of time, the iTAG and WHO agreed that sites would be categorized and analysis focused on sites that met defined performance characteristics to determine whether they achieved the original 2008 objectives. For IB-VPD, sites that reported data during 2010 to 2012 were categorized as: New sites: site began reporting in 2011 or 2012 A Sites: met both of the following criteria

o Reported data in 11 months per year for at least two years during 2010-2012. o Enrolled 100 suspected meningitis cases per year for at least two years during 2010-

2012 (tier 1) or enrolled 500 suspected meningitis/pneumonia/sepsis cases per year for at least two years during 2010-2012 (tier 2).

B Sites: met both of the following criteria o Reported data in 10 months per year for at least two years during 2010-2012. o Enrolled a total of 100 suspected meningitis during 2010-2012 (tier 1) or enrolled

500 suspected meningitis/pneumonia/sepsis cases during 2010-2012 (tier 2). C Sites: Sites that improved in consistency of reporting and case enrolment between 2011

and 2012 but did not meet the criteria of A or B sites. D Sites: All other sites. For RV, analysis focused on sites that reported > 10 months of data during 2011 and 2012 as well as enrolled > 100 cases during those years. The iTAG and WHO collaborated closely during the strategic review via monthly conference calls to agree on the scope of the review and data analysis plan, as well as to assess progress made, and review data that had been collated. Additional ad hoc technical discussions focused on IB-VPD or RV specific issues and data. The results of all methods were reviewed at a 5 day meeting beginning the week of the 16th of September 2013, which brought together the iTAG, WHO and its partners to formulate agreed conclusions and recommendations.

8

Key findings of the 2013 strategic review of the IB-VPD and RV surveillance networks 1. IB-VPD surveillance findings

In 2008, the IB-VPD network comprised 36 reporting countries (69% GAVI-eligible) with 91 sites (60% in GAVI countries) that enrolled 16,124 children with meningitis and 20,098 children pneumonia /sepsis. In 2012 the network comprised 58 reporting countries (79% GAVI-eligible) with 150 sites (70% in GAVI countries) that enrolled 20,098 children with meningitis and 35,480 children with pneumonia and sepsis. Sites1 that received financial support (e.g. sites in GAVI-eligible countries), in general, performed better. Excluding the 37 new sites, 48 (52%) of the 93 GAVI sites conducting meningitis surveillance were categorized as A, B, or C sites while only 10 (20%) of the 51 non-GAVI sites were (Table). Among the 34 category A sites conducting meningitis surveillance, 32 (94%) were located in GAVI-eligible countries. Table: Invasive bacterial vaccine preventable disease surveillance sites conducting meningitis surveillance by category based on data reported to WHO, 2010 to 2012.

Sites Conducting

Meningitis

Surveillance

Category of Site

A

# (%)

B

# (%)

C

# (%)

D

# (%)

Total

# (%)

GAVI sites 32 (94) 14 (64) 2 (100) 45 (52) 93 (65)

Non-GAVI sites 2 ( 6) 8 (36) 0 41 (48) 51 (35)

Total 34 (100) 22 (100) 2 (100) 86 (100) 144 (100)

*The 37 new meningitis sites are excluded; 28 of these sites are in GAVI-eligible countries. The strategic review team agreed that the current IB-VPD network had met the original 2008 objectives for documenting presence of disease (e.g. 61 of 65 countries that reported data during 2008 to 2012 documented presence of pneumococcus), using surveillance as a platform to conduct special studies (e.g. Brazil and Mongolia), and using data to inform vaccine introduction decisions. Capacity for conducting surveillance had been built, not only for meningitis but for other VPDs (such as typhoid detection). Some countries had 2 years baseline data of pneumococcal isolates and 1 year post-PCV introduction data. However, the laboratory network assessment (see Annex) suggested that countries, regions and HQ were not working together as a network. The data generated was not being used in the most efficient and effective manner (see Annex). Because of the need for high-quality data and resource constraints, it was concluded that the investment of financial and technical support should be focused upon a smaller number of sites. 2. RV surveillance findings During 2008-2012, 265 sites in 67 countries reported any data to WHO. Among sites that reported data during 2011-2012, 79 sites (80% GAVI-eligible) in 37 countries (86% GAVI-eligible) met the inclusion criteria. Thirteen of the 37 countries had introduced RV vaccine nationally, and 1 country introduced sub-nationally. Most sites included in the analysis met the targets for the 6 established surveillance performance indicators. Among these sites, the mean RV detection of the 75,353 tested children was 36%, with the largest percentage positive (42%) in the 6-11 month age group. Sixty-five deaths among RV positive cases were reported, with the case fatality ratio ranging from 0% to 5.9% of cases by site. RV seasonality differed by region and by AFR sub-region, with seasonal peaks seen more obviously in AMR, EUR, WPR and the AFR south sub-

1 Year-to-year variability in the number of reporting countries and sites exists due to reasons including in-country

conflict, and political change.

9

Region. Guatemala had sites that met the inclusion criteria and had at least 2 years pre- and two years-post RV vaccine introduction data. In Guatemala, the proportion (mean) of children hospitalized with diarrhoea with RV dropped from 42% in the 2 years pre-vaccine introduction was 42% to 28% in the second year post introduction. The strategic review team agreed that the RV network met the original 2008 objectives for documenting presence of disease, describing disease epidemiology, using surveillance as a platform for special studies in some countries, and using the data for policy decisions. At least 37 (55%) of 67 participating countries had collected at least 2 years of pre-vaccine introduction data that met the inclusion criteria for quality as defined via the strategic review. 3. Assessment of IB-VPD and RV laboratory networks The independent assessment of the laboratory networks found both laboratory networks had made huge advances in strengthening national and regional laboratory capacities, standardization of laboratory testing, improving laboratory quality and performance, and establishing functional international networks. Both, however, remain in a state of development. Of the 2 laboratory networks, the RV network is significantly closer to achieving effective support of the programme. The RV laboratory network has inherent advantages over the IB-VPD laboratory network in that not only is laboratory testing significantly less complex, but laboratory requirements are more compatible with those of existing global networks, including polio and measles/rubella. While WHO has been very effective in providing and contributing to technical guidance of the laboratory networks, provision of overall management and oversight has been less impressive. The fidelity with which laboratory data is entered into the reporting system and the completeness of reporting is clearly inadequate. This problem can be addressed to some extent by ensuring strict linkage between laboratory data and clinical and epidemiology data through the more effective use of unique case ID numbers. This is not a technical problem; it is one of network management and oversight. (full report in Annex). 4. MoH perspectives Standardized questionnaires were received from 27 MoH respondents in AFR, 13 in EMR, and 8 from WPR. Respondents noted the main reasons for establishing surveillance were to assess disease burden, prepare for vaccine introduction, monitor vaccine impact, and to strengthen bacterial laboratory capacity. Overall, 89% of respondents felt surveillance provided information to advocate to, or within, the government on the importance of rotavirus diarrhoea or paediatric bacterial meningitis. National surveillance was reported as the most important factor in the decision to introduce a vaccine or to continue national funding for a vaccine, followed by availability of other national data (e.g. special studies, passive national health systems) and disease burden estimates from WHO or the literature. MoH respondents noted a desire to increase the number of existing sentinel sites in order to obtain more representative data for their country. 5. WHO support to the network To achieve the aim of good quality data, WHO recognized the necessity of a solid infrastructure for surveillance; thus, WHOs activities during the past 5 years have been to develop the building blocks such as instituting laboratory external quality assurance, standardized sentinel site assessments, and training. In addition, WHO provided sites with specimen collection and laboratory supplies including rapid diagnostic kits. Access to advanced laboratory testing was provided via molecular testing at reference laboratories. Regional and global meetings brought together network members to share experiences and foster the feeling of belonging to a network, supported by the dissemination of twice yearly global surveillance bulletins. To improve quality, IB-VPD communication tools (posters and pamphlets) to support specimen collection, identifying bacterial causes of meningitis, and data management were also developed and distributed to sites. A field-based method for estimating the denominator of at-risk children at a sentinel

10

hospital was developed for meningitis surveillance to better understand catchment populations and enable estimation of incidence of meningitis hospitalizations. From 2008 to 2012, GAVI provided $45.5 million in support of both surveillance networks of which $8.8 million (19%) supported WHO staff and $36.7 million (81%) supported surveillance activities. In general, two-thirds of activity funds supported the surveillance infrastructure of the reference laboratory network, regional/global meetings, trainings, technical support, and one-third supported sentinel sites including purchase of supplies. Additional funding provided by the U.S. Centers for Disease Control and Prevention included $530,000 for standardized site assessments, middle income countries and the data landscape analysis, as well as provision of 3 full-time staff and additional technical support. The current year-to-year funding with uncertainty around funding sustainability has caused hesitation by countries and WHO to implement improvements. This uncertainty has been a particular problem for reference laboratories as the year-to-year funding prohibits hiring of staff and long-term planning. This has led to one reference laboratory withdrawing from the network. 6. Data landscape analysis findings The assessment found the data management process is variable by WHO RO, and the data transfer and reconciliation process is cumbersome and time-consuming. Thus, it is warranted to explore new and different options to streamline the current system, while taking into account the unique setting and needs of each WHO RO. The general data flow in each RO is consistent Data capture occurs at a sentinel site, sent to the country office, and then sent to the regional office and then finally submitted to HQ. Two of the ROs have opted to only collect aggregate data while the other 4 collect case-based dataEvery RO has implemented a standard paper case report form, however Each country and even sitemay have made modifications to the standard forms to allow for capture of other variables... Every RO has implemented some type of standard data collection tool, however there is not any standardization across the tools. Some of the major factors affecting the quality of data include relationship with Regional Reference Labs [RRL] with inability to link RRL data [to the individual case data] due to lack of use of unique identifiers and the general availability of resources and funding. A future data management system needs to be flexible and dynamic to account for the different data infrastructures and approaches of each RO. The goal of the new system would to provide a centralized system which would allow for decentralized management and control of data in a case-based format (full report in Annex) Key conclusions and recommendations from the 2013 strategic review 1. IB-VPD surveillance: Conclusions and recommendations The IB-VPD programme is at a critical juncture in its development, and decisive changes will be necessary to produce high-quality data. It was recognized that building a strong and effective IB-VPD surveillance network is both complex and challenging. The overall opinion amongst all partners involved was a strong desire to continue strengthening the network; however, it was agreed that the quality of the network must be urgently improved, and that both technical and financial support should be targeted to a smaller number of sites. In the short-term, these sites would be those that met the defined funding criteria and minimum standards (in general, the A, B, and C category as well as new sites joining the network since 2011). The strategic review team recognized that the category D sites, in general, required investment of considerable human and financial resources before capacity is generated to produce quality data; thus, exclusion of Category D sites would enable focus on the other sites to further improve quality. Furthermore, it was discussed and agreed by ROs, HQ, and partners that collection and sharing of case-based data across all levels is essential to conduct the necessary data analysis to monitor and evaluate the programme, as well as to link clinical and laboratory data particularly of serotype/group data. Data analysis should be conducted more frequently and actively used to monitor surveillance

11

performance and outcome. Strengthening serotype/serogroup data availability and usefulness is also of importance. Recommendations for the IB-VPD Network Size and Scope

Network participation requires meeting minimum data quality standards regarding consistent reporting, enrolment of cases, laboratory, and surveillance performance indicators. In principle, all (GAVI and non-GAVI eligible) countries should be encouraged to participate in the network; however, network size may need to be limited as WHO support to countries to ensure generation of high-quality data is subject to financial and human resource constraints;

In the coming year, the current number of sites in each country should not be expanded unless there is a compelling reason to do so. Since MoH expressed a desire to expand the number of sites, WHO should engage MoH officials in further discussion around this issue;

Only sites that meet the established funding criteria should receive funding in 2014. Sites categorized as D, in general, do not meet these funding criteria, and should not be provided with GAVI funds, if located in GAVI-eligible countries. This decision should be based on data provided during the first 6 to 9 months of 2013, and countries should be notified of the decision by February 2014 (which is prior to receipt of 2014 funding);

During early 2014, additional work is required to clearly define longer-term network objectives, as well as performance and funding criteria. Detection of a minimum number of pneumococcal cases per site per year in countries which have not introduced PCV should be included as a metric. Additionally, sites should be further categorized as: o Fully recognized site: site has been assessed within the past 2 years using the WHO

standardized assessment tool, and is providing quality data for measuring vaccine impact (in pre-PCV introduction countries) or for monitoring serotype epidemiology (post-PCV introduction countries);

o Supported site site receives funding from WHO, agrees to share case-based data on a quarterly basis, and is either a fully recognized site or working towards that;

Networking should be strengthened to share resources and to enhance laboratory collaboration with measles/polio surveillance efforts.

Recommendations for Data Management and Dissemination

Institute a common case-based data system that shares standardized data across sites/regions/reference labs/HQ, with real-time verification and analysis capacity, and with improved data quality;

Strengthen data management capacity for data analysis, interpretation and dissemination at all levels; and

Modify the WHO Global Bulletin to show all reporting countries but limit analysis to a subset of sites reporting quality data, including reporting for at least 12 months and enrolling a minimal number of cases.

Recommendations for the IB-VPD laboratory network:

Conduct in-depth Regional reviews of laboratory networks function and output to identify region-specific issues and provide region-appropriate priority activities;

Reduce the number of participating network laboratories to more closely match programme capacity to fully support and supervise these laboratories to an extent that guarantees confidence in reported laboratory results;

Review and revise the roles and responsibilities of WHO Regional and Global laboratory coordinators to place more emphasis on active management of network performance;

Every effort should be made to assess every laboratory at least once each year; Continue the external quality assurance, and ensure quality control (all positive specimens

and 10% sample of negative specimens should be submitted in a blinded fashion for RRL testing) with data analysis to validate test and laboratory performance;

Report serotype/serogroup data at least twice yearly to WHO HQ and more frequently to ROs

12

to enable more prompt detection of problems with subsequently actions to improve quality; Standardize sample selection for serotyping/grouping; and Link case-based clinical and epidemiological data from sentinel sites to local laboratory results

and polymerase chain reaction results as well as serotype/group data from RRLs, which may require modification of existing data management systems.

Recommendations for sentinel sites

Ensure sites meet a definition of a sentinel site so surveillance objectives can be met; Initiate zero reporting to enable differentiation between no cases enrolled or no report

submitted to WHO; Develop strategies to improve performances at all sites (GAVI and non-GAVI) based on

findings from monitoring and evaluation of the surveillance system; and Prioritize site assessments using external reviewers, MoH, local institutions, etc. 2. RV Surveillance: Conclusions and recommendations

The review team concluded that the RV surveillance network was overall generating quality data useful for decision makers. However, further targeted enhancements to the network would enhance its capacity to provide more standardized information to national, regional, and global policy makers. In particular, all sentinel sites must meet the standard definition for a site so that calculations of rotavirus positivity are based on hospitalized children, rather than children seen at peripheral clinics. Additionally, genotyping data generation should be further strengthened by development of standardized protocols for specimen selection and linking of case-based data that includes clinical and vaccination data. Recommendations for the overall RV surveillance system

Network participation requires meeting minimum data quality standards regarding consistent reporting, enrolment of cases, laboratory, and surveillance performance indicators. In principle, all (GAVI and non-GAVI eligible) countries should be encouraged to participate in the network; however, network size may need to be limited as WHO support to countries to ensure generation of high-quality data is subject to financial and human resource constraints;

The current surveillance performance indicators should be examined for usefulness; and Encourage further networking including laboratory collaboration with measles, polio, IB-VPD

surveillance and sharing of resources. Recommendations for Data Management and Dissemination

Countries should report case-based data to all regional offices and HQ, and gathered data should be further standardized;

RRLs should report genotype data to ROs twice yearly to ROs, as feasible within current resource constraints; and

Global Bulletins should be modified to show all reporting countries on a map, but limit analysis to subset of reporting sites that report every month and enrol > 100 cases annually.

Recommendations for the RV laboratory network

Standardize sample selection for genotyping; Begin to establish the linkage of case-based data to genotype data; Examine country-level genotype distribution in addition to distribution at regional and global

levels; Build additional technical capacity at the national laboratory (NL) level, if possible; and Funding permitting, include all sites in the EQA programme. Recommendations for sentinel sites

Ensure sites meet a definition of a sentinel site so surveillance objectives can be met; Update eligibility criteria for inclusion in data analyses to include 12 months reporting and

>100 specimens at a single site (no satellite sites)

13

Initiate zero reporting to enable differentiation between no cases enrolled or no report submitted to WHO; and

Develop strategies to improve performances at all sites (GAVI and non-GAVI). Strategies may be based on findings from monitoring and evaluation of the surveillance system.

3. IB-VPD and RV future vision and complementary approaches for IB-VPD surveillance: Conclusions and recommendations

Recommendations for complementary approaches to IB-VPD surveillance For IB-VPD, there is an urgent need to improve the quality of the data, and to monitor ongoing serotype epidemiology. However, it is anticipated that data quality will be unable to quickly reach the desired level in all countries, particularly those with the weakest health and hospital systems. Thus, a complementary strategy to IB-VPD surveillance for measuring vaccine impact is required. Develop tools and processes (models) that allow PCV impact assessment by input of quality

data from IB-VPD as well as complementary existing data sources including pneumonia morbidity and mortality.

Recommendations for assessing future utility for surveillance of other VPDs Both the IB-VPD and RV sentinel hospital surveillance networks have developed a surveillance platform of improved epidemiological and laboratory capacity. Careful consideration is warranted to determine whether other VPD surveillance activities can be added to this platform, without jeopardizing the quality of IB-VPD and RV data. Define how to best build upon these networks to conduct surveillance for other VPDs,

including those prevented by current vaccines and future vaccines such as typhoid, as well to increase capacity for outbreak detection.

Appropriate surveillance approach(es) for other VPDs should be developed based on the specific disease characteristics (incidence, hospitalization, diagnostic testing characteristics, etc.) and the key objectives for surveillance of that disease.

Recommendations for monitoring implementation of the strategic reviews recommendations The iTAG should continue to engage in the on-going progress made from implementation of

these recommendations via quarterly teleconferences or meetings, and advice WHO on any corrective actions that may be required.

Recommendations for funding Improvement of the data management system with a move to case-based data (including

linking of data from reference laboratories) and more on-going data analysis, interpretation and dissemination will require additional financial support. Longer term funding would enhance commitment to make changes in the system.

14

The Informal Technical Advisory Group for New Vaccines Surveillance Assessment of the

2013 Strategic Review of the Invasive Bacterial Vaccine Preventable Diseases and Rotavirus Sentinel Surveillance Networks

Introduction and background:

The informal Technical Advisory Group (iTAG) would like to acknowledge the foresight of WHO in

requesting a strategic review of the Rotavirus (RV) and Invasive Bacterial-Vaccine Preventable

Diseases (IB-VPD) Surveillance Networks. As WHO and partners consider how the next phase of

surveillance for RV and IB-VPD should be designed, reviewing the accomplishments and

challenges during the first five years of network development has provided valuable insights into

approaches that have been successful and identified areas of needed improvement. The

accompanying Executive Summary and report provide the detailed scope, process, and results of

the strategic review. This document highlights the iTAGs identification of several overarching

issues and policy considerations.

The RV and IB-VPD networks have accomplished the 2008 objective of documenting the

presence of disease; the resultant data have contributed to decisions of many countries to

introduce RV vaccine and Pneumococcal Conjugate Vaccine (PCV). As a result, strategic

surveillance objectives for many countries have changed from making the decision to introduce

vaccine(s) to monitoring the impact(s) of vaccine introduction and obtaining data to support

eventual country vaccine financing. In addition, the RV network has successfully initiated

monitoring of rotavirus genotype distribution in all regions, and has promising results for the

potential ability of the network to monitor impact of vaccine introduction. The laboratory capacity

built for IB-VPD in some sites has detected disease due to other etiologic agents, as well as

outbreaks of meningitis. Pneumococcal serotype information has been collected from previously

under-represented countries and regions. The ability of most IB-VPD sites or network to serve as

a platform to monitor impact of PCV introduction remains to be demonstrated.

The review process has confirmed that the capacity needed for a site to successfully implement

RV surveillance is quite different from that needed for IB-VPD surveillance. Gastroenteritis is a

common, easily recognized clinical condition, and sample collection to test for RV is non-invasive.

IB-VPD Tier 1 sites are conducting surveillance for meningitis, and bacterial meningitis at a

population level is an infrequent albeit severe condition, requiring a substantial population per site

in order to detect more than a small number of cases. For the IBD-VPD Tier 2 sites which are

also conducting surveillance for bacteremia, sepsis and pneumonia, case confirmation is

challenging. Blood cultures are not routinely performed in many countries and widespread use of

antibiotics prior to seeking medical carecoupled with poor blood culturing techniqueshinder

isolation of bacteria. Also, bacteremia occurs in a small percentage of pneumonia cases, and

diagnostic testing for non-bacteremic pneumococcal pneumonia is neither sufficiently sensitive

nor specific with present assays. For both IB-VPD tiers, establishing laboratory capacity to isolate

bacteria from normally sterile body fluids consistently and reliably remains a challenge in many

sites; as Ray Sanders, the consultant who reviewed the network laboratory activities notes, the

laboratory testing required for the RV network is significantly less complex".

Since 2008, when the IBD-VP network development began, the field of PCV impact evaluation

has gained substantially more experience in monitoring the changes in hospitalized pediatric

pneumonia (i.e., impact on a syndromic outcome that is not specific for pneumococcal

15

pneumonia). It may now be possible to monitor PCV impact on hospitalized pediatric pneumonia

in selected sites, to complement the original focus on Invasive Pneumococcal Disease as an

outcome. However impact monitoring by using a pneumonia syndrome endpoint is also complex

and feasible in only selected settings. RV surveillance for monitoring vaccine impact may also

require special studies on vaccine effectiveness and intussusception, which may not be feasible

in all sites.

In order to support country decision making for vaccine introduction and for sustaining vaccine

programs, countries need to be able to estimate vaccine impact. New tools for such assessments

are becoming available, including increased experience with models in which countries can use

either their own data, data from special studies, or global burden of disease data to estimate the

impact of introducing pneumococcal or rotavirus vaccine on their countrys morbidity and

mortality, as well as the cost-effectiveness of vaccination. Surveillance networks may need to

consider how data generated through the network and other data sources could be used as

inputs into such models, in place of monitoring actual disease impact data at the country level.

The iTAG review identified certain areas where modifications would substantially enhance the

utility of the data and the allocation of resources.

Issues identified during the review: 1) As noted in both the laboratory review and the data landscape review, the system has

developed and tends to function as six separate regional networks, rather than as a single

network. Implementation of the IBD-VPD and RV networks has largely been designed and

managed at the regional level. Some degree of regional flexibility is appropriate and

necessary to permit accommodation of the surveillance to regional and national level needs

and realities. However the network also needs overall project management and

accountability, and the ability to undertake global assessment and analyses.

2) While the analyses of the system in the accompanying report provide useful information

about the networks, additional analyses of the IB-VPD data at a global level that were

proposed as part of the review were not feasible in the available time frame. Issues limiting

analyses included the lack of case level data in an agreed upon standard format, as well as

difficulty linking specimen laboratory results from some laboratories with the clinical and

epidemiologic data on the same patient. This was also an issue with data collection with the

RV network, although not as critical. The ability within the network to do real time

performance monitoring was therefore limited.

3) The resources to support a network of this size at the country, regional and global level are

insufficient and could benefit from more strategic allocation. Two specific examples were

noted by the iTAG during the review. At the global level, there were inadequate resources

for management, and data processing and analysis (exacerbated by the data system issues

noted elsewhere in this document). The review also noted substantial differences in the

regional staffing and management of the network(s). For example, the number of

sites/countries supported by each Regional Office and Regional Reference Laboratory varied

widely. In a region with a large number of sites, the regional staffing is unlikely to be

sufficient to support the recommended on-site assessments of either the surveillance

program or the site laboratories.

In addition the decreased funding available in 2014 increases the urgency of decisions about appropriate resource utilization and sizing the network relative to the resources.

16

Representatives from the Regional Offices also note that the current approach of year-to-year funding has made it difficult to undertake longer-term planning and investment at all levels of the networks.

4) For the IB-VPD network, preliminary data suggest a wide range in the number of laboratory

confirmed cases per site, with some sites identifying small numbers of cases, and showing

substantial year-to-year variability.

5) The value of the networks for countries and regions includes the technical standards for

surveillance that are established (which can serve as models for other countries to follow);

the access to technical assistance, reference laboratories, and reagents; and the

communication and collaboration that occur among participant countries and sites, including

countries which are not GAVI eligible, but may be encouraged to undertake surveillance and

introduction of vaccines.

Recommendations: 1) The RV and IB-VPD networks should build on the experience and results of the past five

years to refine appropriate objectives and strategies to meet future surveillance needs.

A) For RV, the network appears capable of meeting the primary objectives, monitoring the

impact of vaccine introduction and monitoring changes in genotype distribution. It is

important to define the minimum data required to meet surveillance objectives at the

national and regional levels. Specific modifications for the RV network proposed in the

review report will strengthen the networks ability to meet these objectives.

B) For IB-VPD, the network should refine the objective and performance criteria for

surveillance for Invasive Bacterial Disease, incorporating what has been learned from the

initial experience. An updated objective should include the requirement that participating

sites demonstrate the ability to document a substantial number of IPD cases, in order to

establish a credible baseline of IPD for at least two years before vaccine introduction. A

performance criteria of an average of 20 to 30 cases of invasive pneumococcal disease

each year before vaccine introduction is proposed, based on the analysis of sentinel site

comparability to population level trends in the US active surveillance systems. This target

also reflects the concern that sites with smaller numbers are likely to exhibit greater year

to year variability, increasing the potential for misleading interpretations of secular trend

data, especially regarding non-vaccine serotypes. Relatively few sites may be able to

identify this number of IPD cases annually in the target age range. However, focusing the

surveillance resources on a smaller number of sites with more interpretable data may

provide a better model of what can be accomplished by surveillance, and encourage

countries considering vaccine introduction to invest country or donor resources in

surveillance (see below).

C) For countries and sites that are not currently able to meet the above performance criteria

for IBD/IPD, other approaches to the core objectives should be explored.

1. To address the objective of monitoring vaccine impact, WHO could explore the

use of a hospitalized pediatric pneumonia endpoint in a limited number of sites

where population size and available data sources suggest that such an

17

approach might be feasible, building on the conclusions from the PCV impact

meeting.

2. To meet country needs for decision making on vaccine financing, WHO could explore encouraging the use of existing or future modeling approaches to estimate the impact of introducing pneumococcal vaccine on country specific morbidity and mortality, as well as the cost-effectiveness of vaccination. 3. Monitoring the impact of pneumococcal vaccine on NP colonization, especially for vaccine serotype strains, has also been proposed as an alternate method for countries to bridge to disease impact data from similar settings, using WHO recommended methodologies for nasopharyngeal swabbing. WHO should consider the conclusions from the PCV impact meeting regarding the possible utility of this approach; concerns include the technical requirements for undertaking NP colonization surveys, the costs for such surveys, and the complexity of extrapolating from NP colonization data to disease impact especially for non-vaccine serotype disease.

D) An additional objective for IB-VPD surveillance is to support improved bacteriologic

capacity to identify other important pathogens and outbreaks. Monitoring of anti-microbial

susceptibility also remains an important question. However, the experience of the past

five years suggests that either these objectives should be achieved by special studies in a

limited number of well-resourced centers, or there needs to be a substantial improvement

in sentinel site and laboratory performance, with increased technical assistance, site

assessments, and close to real time performance monitoring.

The iTAG encourages efforts to further develop IB-VPD surveillance capacity to meet these objectives, and also to support sites that seek to eventually meet the performance criteria proposed in B above. WHO should encourage countries to explore strategies for supporting this capacity building, through alternative (or supplemental) funding such as GAVIs Health Systems Strengthening, in-country support for VPD surveillance, and other sources. However the iTAG also notes that site funding is not the only needed input; the ability to support the network with access to technical assistance, reference laboratories, reagents, site assessments, data reporting systems, communication and collaboration also require substantial resources which are currently limited.

2) It is essential for the future that RV and IB-VPD surveillance each function as single global

networks generating credible well-defined data. The network should also facilitate efficient

use of the data collected for core network functions such as monitoring system performance in

real time, contributing data for laboratory quality assurance, and evaluating the performance

of new diagnostic methods. This will require modification of data systems and implementation

of policies which facilitate relevant data collection and timely analysis.

Specific needs include:

A) Use of a standard approach for variable names and coding B) Use of unique identification numbers for patients and/or specimens, policies which

ensure that laboratory results (site and RRL) are linked with clinical and epidemiologic data on each specimen/patient, and policies that assure an appropriate sample of specimens are tested in Regional Reference Laboratories

C) With appropriate security safeguards, sharing of case level records among all levels of the system

D) System flexibility to incorporate new laboratory tests when officially added as standard procedure

18

E) Zero or negative reporting from sites, so that absence of cases or variables can be differentiated from lack of reporting

F) Software with editing and verification capability to improve data quality Since it is unlikely that the time or funds will be available for development of entirely new data systems, the iTAG recommends a rapid evaluation of the feasibility of modifying existing web-based systems used within the networks to meet the above system requirements. The process of the strategic review itself has resulted in substantial work to map the variable names across sites and to understand what the values for variables mean from site to site, which should contribute to progress toward a single network. The need for stronger networking, communication, and policies at all levels cannot be over-emphasized.

3) There should be better targeting of resources linked to the refined objectives.

The iTAG commends the thoughtful effort by the IB-VPD network to define performance criteria for joining the network and for continuing to participate as a network site. These criteria have also been used to identify sites which have not met the criteria over the past two years. Given the decreased resources available for 2014, the iTAG concurs with focusing resources on those sites that are able to meet the performance criteria and eliminating funding for the sites that have not. Looking beyond 2014, funding and human resources should be more closely aligned with the refined objectives and strategies above. While the iTAG agrees the funding cycle should permit more than year-to-year funding, especially when staff must be hired, it is also critical that continuation of funding be contingent on sites meeting meaningful performance criteria. There should be performance agreements for all participants in the network, including the surveillance sites, regional offices, regional reference laboratories, global reference laboratories, and WHO HQ, specifying in detail responsibilities as well as resources to be received. Estimates for the approximate number of site visits to be made and specimens to be processed would contribute to a more realistic estimate of the number of sites that could be supported given limited human and financial resources. Available resources should be allocated to assure data capable of meeting system objectives; the number of sites that can be supported is finite. The iTAG recommends that WHO implement structured performance-based agreements with all participants. Sites/countries could consider the refined list of objectives and identify, based on their experience to date, which approach for vaccine impact monitoring (e.g. IPD, hospitalized pneumonia, modeling, etc.) they would be best suited to undertake. Funding for reference laboratories could be linked to best estimates of the number of site visits to be done and the number of specimens to be processed.

While capacity building is a goal of the RV and IBVPD surveillance programs, regular monitoring and accountability are needed to ensure a level of performance that provides data to meet country and region specific needs. The network provides technical specifications and support; it also provides funding to country sites to establish successful models. Other countries can then build on these models, using country funds or other sources such as GAVI HSS funds. As countries introduce vaccines they are committing themselves to large investments in the future for vaccine procurement; surveillance costs are small compared with vaccine purchase costs but are essential to vaccine sustainability. The vision for the networks is that countries have models for disease surveillance that enable countries and regions to build credible, high quality surveillance systems which will provide reliable data on vaccine preventable diseases.

19

1. BACKGROUND

The Global Invasive Bacterial-Vaccine Preventable Diseases and Rotavirus Surveillance

Networks

Prior to 2008, surveillance for invasive bacterial diseases and rotavirus disease primarily was

conducted by special projects and regional surveillance networks that had been established

through funding from the GAVI Alliance via the Accelerated Development and Introduction Plans

(ADIPs) and the Hib Initiative. Given time limited funding through the ADIPs and the Hib Initiative,

global immunization partners and Ministries of Health supported the transition of these special

projects and regional surveillance networks to the formation of global invasive bacterial-vaccine

preventable diseases (IB-VPD) and rotavirus (RV) surveillance networks that would operate

under the auspices of the World Health Organization (WHO).

Since 2008, with support from the GAVI Alliance and coordination by WHO, the global IB-VPD

and RV surveillance networks have operated under similar management structures, which include

networks of sentinel surveillance sites and laboratories that enroll cases and conduct initial

testing. Sentinel surveillance sites report data either to the local Ministry of Health, WHO Country

Office or the WHO RO. The laboratory networks are structured under a tiered model in which

sentinel site laboratories are support by National Laboratories (NLs) and/or Regional Reference

Laboratories (RRLs); NLs are support by RRLs, and RRLs are supported by the Global

Reference Laboratories (GRLs). Both clinical and laboratory data are reported to WHO HQ on

semi-annual or annual basis.

From 2008-2012 under this new management structure, WHO implemented strategies to improve

data quality, including provision of technical support, standardized site assessments, and

purchase of supplies and equipment for sentinel sites. An overall timeline of key events includes:

2009-2009: transition of the networks to WHO coordination with GAVI financial support, data standardization meeting held with agreement on collection and sharing of data, culminating in the dissemination of the first global IB-VPD and RV surveillance bulletins in December 2009;

2010: RV standardized sentinel site assessment tool developed, first global surveillance meeting convened with agreement on standard meningitis case definitions, and revision of IB-VPD surveillance performance criteria;

2011: visual aid standard operating procedures begun for IB-VPD (posters for collection of cerebrospinal fluid and laboratory identification of bacterial from CSF as well as a data management pamphlet), IB-VPD laboratory external quality assurance programme piloted, second global surveillance meeting held with agreement to strengthen identification of bacterial pathogens by addition of rapid diagnostic tests and PCR at RRLs, and agreement for sharing data on all suspect meningitis cases identified rather than limited to cases classified as probable bacterial meningitis;

2012: IB-VPD standardize assessment tools finalized; IB-VPD EQA launched for all participating laboratories; IB-VPD visual aids finalized; manual for laboratory detection of CSF bacterial pathogens finalized; first RV and IB-VPD laboratory technical working group meetings held; RV EQA undertaken for RRLs; 3rd global surveillance meeting held;

2013: strategic review of IB-VPD and RV surveillance networks launched with review of experiences from 2008 to 2012.

During 2008 to 2012, both IB-VPD and RV surveillance networks expanded (Figures). In 2008,

the IB-VPD network comprised 36 countries reporting sentinel site data (with one additional

country reporting laboratory based data), of which 69% were GAVI-eligible countries. In total, 91

sentinel sites reported data with 16,124 meningitis cases and 20,098 pneumonia/sepsis cases

reported. In 2012, 58 countries reported data, 79% GAVI-eligible, via 150 sentinel sites. In total,

20

20,098 meningitis cases and 35,480 pneumonia/sepsis cases were reported. For RV, the

number of reporting countries increased from 36 (56% GAVI eligible) in 2008 to 60 (72% GAVI-

eligible) in 2012. The number of sites increased from 131 (40% in GAVI-eligible countries) to 178

(55% in GAVI-eligible countries), and the number of children

21

Figure 3. Map of Sentinel Sites in the WHO RV Surveillance Network in 2008

Figure 4. Map of Sentinel Sites in the WHO RV Surveillance Network in 2012

22

In 2008, the following overarching objectives were established for the global IB-VPD and RV

surveillance networks:

1. To generate local data for decision making and informing policy regarding Haemophilus

influenzae type b (Hib) vaccine, pneumococcal conjugate vaccine (PCV), and rotavirus

vaccine introduction and sustained use

2. To assess and monitor disease trends and serotype/genotype distribution over time

(nationally, regionally, globally)

3. To develop a platform for vaccine impact studies

4. To highlight the value of surveillance data and use for fundraising and advocacy

More specific objectives included the following, according to pre- and post-vaccine introduction

periods:

Time period Objectives

Pre-vaccine introduction period

To document presence of disease,

describe the disease epidemiology and

provide data for estimating disease

burden

To establish systems to measure impact

after vaccine introduction

To identify circulating serotypes and

measure serotype distribution

To monitor antibiotic sensitivity (IB-VPD

only)

Post-vaccine introduction period

To assess disease trends over time

To monitor vaccination programme impact

To monitor changes in circulating

strains/serotypes

To use the surveillance networks as

platforms for vaccine effectiveness and

safety evaluations

To meet these objectives, the WHO-coordinated IB-VPD and RV surveillance networks have

worked to build infrastructure in countries and regions to provide quality data to assist with the

planning of public health programmes. These global surveillance networks also include sentinel

sites within countries not eligible for GAVI support that have chosen to provide data voluntarily;

many of these sites are located within WHOs Region of the Americas of the World Health

Organization (AMR) and the Eastern Mediterranean Region.

Strategic review of the IB-VPD and RV surveillance networks

Many elements are needed to build a strong surveillance infrastructure, such as training,

laboratory support, data management, and communication, and although much progress has

been made in strengthening these developing surveillance networks, more can be done. WHO

and global partners determined that it would be timely to conduct an in-depth, five-year strategic

review of the IB-VPD and RV surveillance networks to identify strengths and weaknesses and

23

decide on strategies and actions to ensure that the networks are responsive to the information

needs of all immunization stakeholders.

For this strategic review, WHOs informal technical advisory group of experts (iTAG) for new

vaccines surveillance provided guidance and technical direction (Appendix A). WHO and the

iTAG agreed that the strategic review would have the following objectives:

1. To assess whether and to what extent the original 2008 objectives for the IB-VPD and RV

surveillance networks have been met

2. To critically assess the current IB-VPD and RV surveillance networks, to specifically include:

Ministry of Health (MoH) perspective;

IB-RV laboratory networks;

Capacity building conducted in support of developing and maintaining the IB-VPD and

RV networks; and

Data management, production, dissemination, and use by the IB-VPD and RV

networks

3. To provide conclusions and recommendations for the future vision for the networks, including

those around enhancing country ownership, laboratory networks, capacity building and data

management, particularly:

Clarify what should be the current objectives for both networks including contributing to

vaccine impact assessments (short- and long-term);

Determine optimal strategies to meet these objectives and overcome identified

common challenges;

Define key indicators/milestones to monitor progress; and

Define potential complementary approaches for IB-VPD surveillance

To accomplish this, beginning in March 2013, the iTAG met via monthly teleconferences to

discuss the scope of the review, data analysis plans and, later on, results from the various

components of the review. A 5-day global meeting on the strategic review was convened 16th

September 2013, during which findings were presented and discussed, and country specific

examples of IB-VPD and RV surveillance systems and surveillance data use were presented by

representatives from each of the WHO ROs. Final recommendations from the strategic review

process are provided in this report.

2. IB-VPD SURVEILLANCE NETWORK DATA ANALYSES

Analysis of Data Reported to WHO

METHODS

Source of surveillance data

In the second quarter of 2013, WHO Headquarters (HQ) requested that the Regional Offices

(ROs) submit case-based sentinel site-level meningitis (tier 1), pneumonia (tier 2) and population-

based (tier 3) data. Analysis of IB-VPD surveillance data was conducted on all data available to

WHO, including case-based data collected by 4 ROs and data aggregated by sentinel site in 2

ROs. Data from over 70 databases were consolidated and cleaned. Due to variability in the data

24

reported to WHO HQ, the data from each region was analyzed separately and one region

required separate analyses by country. IB-VPD variables were mapped across Regions, and the

first uniform data analysis performed. Due to the volume of data and paucity of time, the iTAG

and WHO agreed that sites would be categorized and analysis focused on sites that met defined

performance criteria to determine if they achieved the original 2008 objectives (refer to data

analysis plan in Annex)

Components of the surveillance data analysis

The below Table lists the components of the data analysis performed as part of the strategic

review. All sites were assessed for whether the presence of disease was documented. The

additional components of the analysis were limited to A category sites.

Table 1. Summary of data analysis components for the 2013 IBD surveillance network strategic review.

Category Data analysis components

Document presence of disease Number and percent of countries / sites that

documented the presence of the VPD bacterial pathogens

Surveillance performance indicators and

data completeness

Percent of sites that report according to the agreed timeline

Percent of suspected meningitis cases that have a lumbar puncture (LP) performed

Percent of meningitis cases with age data available Percent of lumbar punctures performed that have a

culture result recorded

Describe disease epidemiology and

provide data for estimating disease burden

Percent age distribution of suspected meningitis cases

Percent probable bacterial meningitis Percent age distribution of probable bacterial

meningitis Percent age distribution of lab-confirmed cases (Hi,

Spn, Nm) Percent probable bacterial meningitis due to Hi,

Spn, Nm organisms Percent lab-confirmed cases from suspected

meningitis cases with an LP Percent culture-positive suspected meningitis

cases with VPD organisms detected by culture Percent Binax positive for Spn from suspected

meningitis cases with an LP Percent Latex positive for Hi, Spn, Nm organisms

from suspected meningitis cases with an LP Percent PCR positive for Hi, Spn, Nm organisms

from suspected meningitis cases with an LP Percent of suspected meningitis cases of known

outcome that died (CFR)

Assess disease trends over time Monthly distribution of suspect and probable

meningitis cases, by year Monthly distribution of lab-confirmed cases, by year

25

Document presence of disease

During 2008 to 2012, all but 4 participating network countries identified Spn cases via IB-VPD

surveillance. The figures below also display the number of cases identified by each country as

well as the year of PCV introduction. Despite most countries having introduced Hib vaccine

during 2008 to 2012, over half the countries were able to identify Hi cases. The final figure shows

identification of Nm.

Figure 5 Numbers of countries and sites participating in the WHO-coordinated IB-VPD surveillance network that identified

Streptococcus pneuomoniae, 2008-2012

Figure 6 Map of countries and sites participating in the WHO-coordinated IB-VPD surveillance network that identified

Streptococcus pneuomoniae, 2008-2012

26

Figure 7 Numbers of countries and sites participating in the WHO-coordinated IB-VPD surveillance network that identified

Haemophilus influenzae, 2008-2012

Figure 8 Map of countries and sites participating in the WHO-coordinated IB-VPD surveillance network that identified

Haemophilus influenzae, 2008-2012

Figure 9 Numbers of countries and sites participating in the WHO-coordinated IB-VPD surveillance network that identified

Neisseria meningitidis, 2008-2012

27

Determination of inclusion sites

The strategic review focused on sites most likely to have accomplished the original 2008

surveillance tasks most completely, sites were thus categorized into performance levels A, B,

C,D or new based on the following criteria.

New: Sites new to the surveillance network in 2011 or 2012.

A Sites: met both criteria

o Reported 11 months of data per year for at least two years during 2010-2012.

o Enrolled 100 suspected meningitis cases per year for at least two years during

2010-2012 (tier 1) or enrolled 500 suspected meningitis/pneumonia/sepsis cases

per year for at least two years during 2010-2012 (tier 2).

B Sites: met both criteria

o Reported 10 months of data per year for at least two years during 2010-2012.

o Enrolled a total of 100 suspected meningitis during 2010-2012 (tier 1) or enrolled

a total of 500 suspected meningitis/pneumonia/sepsis cases during 2010-2012

(tier 2).

C Sites: Sites that improved in consistency of reporting and case enrolment between 2011

and 2012 but did not meet the criteria of A or B sites.

D Sites: All other sites.

Only data from sites that met the A criteria were analysed in detail.

Figure 10. Categorization of sites reporting meningitis data to WHO, 2010-2012

Region

A:

Strict

B:

Moderate C: Improved

D:

Other New

Total

# % # % # % # % # %

AFR 17 35% 6 12% 2 4% 8 16% 16 33% 49

AMR* 0 0% 9 19% 0 0% 28 58% 11 23% 48

EMR** 11 26% 5 12% 0 0% 25 58% 2 5% 43

EUR 0 0% 2 17% 0 0% 3 25% 7 58% 12

SEAR*** 5 71% 0 0% 0 0% 2 29% 0 0% 7

WPR 1 5% 0 0% 0 0% 20 91% 1 5% 22

Total 34 19% 22 12% 2 1% 86 48% 37 20% 181

28

Figure 12. Categorization of Sites Reporting Pneumonia Data to WHO, 2010-2012

Region

A:

Strict

B:

Moderate C: Improved

D:

Other New

Total

# % # % # % # % # %

AFR

AMR* 11 23% 9 19% 0 0% 17 35% 11 23% 48

EMR** 3 50% 0 0% 0 0% 3 50% 0 0% 6

EUR

SEAR*** 4 57% 1 14% 0 0% 2 29% 0 0% 7

WPR 3 50% 2 33% 0 0% 1 17% 0 0% 6

Total 21 31% 12 18% 0 0% 23 34% 11 16% 67

Figure 13. Map of Pneumonia Sentinel Sites Categorized by Levels A, B, C, D, and New

Figure 11. Map of Meningitis Sentinel Sites Categorized by Levels A, B, C, D, and New, 2010-2012

29

Data analyses

All analyses were stratified by region and GAVI funding eligibility. As per surveillance network

protocol, any country eligible for GAVI support at the start of the WHO-coordinated IBD

surveillance network has remained eligible for GAVI support; this includes countries that have

graduated from GAVI eligibility. The data analyses were done for all sites that met the

aforementioned level A inclusion criteria.

Performance indicators. The performance indicators calculated were the following:

Performance indicator Target

Percent of sites that report according to the agreed timeline 80%

Percent of suspected meningitis cases that have a lumbar puncture (LP)

performed 90%

Percent of meningitis cases with age data available

Percent of lumbar punctures performed that have a culture result recorded 90%

Percent age distribution of suspected meningitis cases. The percent age distribution of all

suspected meningitis cases was calculated using data from all cases enrolled at a site from 2008-

2012. The age stratification for this and subsequent distributions are:

Region(s) Age groups reported

30

AFR, EMR, EUR, WPR

0-5 months

6-11 months

12-23 months

24-59 months

AMR

0-11 months

12-23 months

24-59 months

All 59 months

Percent probable bacterial meningitis: The percent of all suspected meningitis cases that had

probable bacterial meningitis was calculated using all available data from a site for 2008-2012.

Only suspected meningitis cases that had data to calculate probable bacterial meningitis status

were used as the denominator.

Percent age distribution of probable bacterial meningitis: The percent age distribution of cases

classified as probable bacterial meningitis was calculated using data from all cases enrolled at a

site from 2008-2012.

Percent age distribution of lab-confirmed Haemophilus influenzae (HI), Streptococcus pneumonia

(Spn), and Neisseria meningitides (Nm): The percent age distribution of all suspected meningitis

cases with a lab-confirmed diagnosis (Hi, Spn, Nm) was calculated using data from all cases

enrolled at a site from 2008-2012.

Percent probable bacterial meningitis due to Hi, Spn, Nm organisms: The percent of all probable

bacterial meningitis cases due to each pathogen was calculated for each year, for 2008-2012.

Percent lab-confirmed cases from suspected meningitis cases with an LP: The percent of all

suspect bacterial meningitis cases due to any lab-confirmed pathogen was calculated for each

year, for 2008-2012. Lab-confirmation was by any method (culture, latex, Binax, PCR).

Percent culture-positive suspected meningitis cases with VPD organisms detected by culture: The

percent of cases with a positive culture for Hi, Spn or Nm was calculated from the total number of

cases tested with culture.

Percent Binax positive for Spn from suspected meningitis cases with an LP: The percent of all

suspected meningitis cases that test positive for Spn by Binax was calculated from the total

number of cases tested with Binax.

31

Percent Latex positive for Hi, Spn, Nm organisms from suspected meningitis cases with an LP:

The percent of suspected meningitis cases that tested positive for Hi, Spn or Nm by latex

agglutination was calculated from the total number of cases tested by Latex.

Percent PCR positive for Hi, Spn, Nm organisms from suspected meningitis cases with an LP:

The percent of suspected meningitis cases that tested positive for Hi, Spn or Nm by PCR was

calculated from the total number of cases that were tested by PCR.

Percent of suspected meningitis cases of known outcome that died (case fatality rate): The case

fatality rate of suspected meningitis cases was calculated from the total number of suspected

meningitis cases that had a known outcome recorded.

Monthly distribution of suspect and probable meningitis cases, by year: The monthly distribution

of probable meningitis cases was analyzed overall and for the years 2008-2012.

Monthly distribution of lab-confirmed cases, by year: The monthly distribution of suspected

meningitis cases with a lab-confirmed diagnosis was analyzed overall and for the years 2008-

2012.

MAIN RESULTS

Sentinel site reporting and inclusion criteria

During 2008-2012, 181 sites reported data to WHO HQ as per the reporting requirements for the

2013 Strategic Review. After applying the inclusion criteria mentioned before, there were 72

sentinel sites that enrolled cases in at least 11 months per year for two years from 2010-2012 and

of those, there were 34 sites that also enrolled more than 100 suspected meningitis cases per

year in two years from 2010-2012, thereby meeting the A site criteria.

Table 2. Sentinel sites meeting reporting and testing criteria, 2010-2012

Region Total # of Sites

# (%) sites that enrolled cases > 11 months per

year for at least 2 years during

2010-2012

# (%) sites that enrolled >100

suspected meningitis cases per year at least 2 years

during 2010-2012

# (%) of sites that meet

both criteria (A

designation)

AFR 49 19 (39) 25 (51) 17 (35)

AMR* 48 23 (48) --- ---

EMR** 43 15 (35) 11 (26) 11 (26)

EUR 12 2 (17) --- ---

SEAR 7 5 (71) 6 (86) 5 (71)

WPR 22 8 (36) 1 (5) 1 (5)

Total 181 72 (40) 43 (24) 34 (19)

*Brazil is excluded (conducts laboratory-based surveillance);

**Morocco: country included but 88 reporting sites excluded (conducts national Tier 3

surveillance)

32

***Bangladesh: 1 population-based field site is excluded

Table 3. Percent of suspected meningitis cases that have a lumbar puncture (LP) performed amongst A-level sites

Region # of sites that meet

criteria

Total # of suspected meningitis

cases

# of suspected meningitis cases that had an LP performed

mean % of suspected meningitis cases that had an LP performed

median % of suspected meningitis cases that had an LP performed

range % of suspected meningitis cases that had an LP performed

AFR 17 19,294 19,150 100 100 (100, 100)

AMR --- --- --- --- --- ---

EMR 11 9,414 7,323 78 88 (19, 100)

EUR --- --- --- --- --- ---

SEAR 5 10,783 6,455 60 66 (17, 93)

WPR 1 268 268 100 --- ---

Total 34 39,759 33,196 84 100 (17, 100)

Surveillance performance indicators and data completeness

The 34 sentinel sites that met both criteria during the period of 2010-2012 were further evaluated

by two of the key meningitis surveillance performance indicators the percentage of suspected

meningitis cases that had a lumbar puncture (LP) performed (target: at least 90%) (Table 2), as

well as the percentage of LP performed that had a culture result recorded (target: at least 90%)

(Table 3). In addition, the completeness of data reported by these sites was assessed by the

availability of age data (Table 4). Age data was not available in some regions because of the

absence of case-based data.

Table 4. Percent of meningitis cases with age data available

Region # of sites that meet criteria

# of cases in

database

# of cases with age

data available

mean % cases with age data available

median % cases with age data available

range % cases with age data available

AFR 17 31,771 31,227 98 100 (81, 100)

AMR --- --- --- --- --- ---

EMR 11 16,216 16,216 100 100 (100, 100)

EUR --- --- --- --- --- ---

SEAR 5 --- --- --- --- ---

WPR 1 300 300 100 100 (100, 100)

Total 34 48,287 47,743 99 100 (81, 100)

33

Table 5. Percent of lumbar punctures performed that have a culture result recorded

Describe disease epidemiology and provide data for estimating disease burden

Percent age distribution of suspected meningitis cases, 2010-2012

The age distribution of suspected meningitis cases was analyzed across regions. Only

information from sentinel sites where the age of suspected meningitis cases was provided during

the period 2010-2012 was used. The distributions vary widely between and within regions.

Disease typically occurs in children too young to be vaccinated which is represented in the WPR

age distribution but only in some sites in AFR and EMR.

Figure 14. Percent age distribution in AFR for 2010-2012

0102030405060708090

100

6 (C

MR

)

32

(CO

D)

10

(ET

H)

14

(GH

A)

15

(GH

A)

16

(KE

N)

19

(MLI)

23

(NA

M)

28

(NE

R)

26

(NG

A)

36

(SE

N)

43

(TG

O)

45

(UG

A)

46

(UG

A)

47

(UG

A)

48

(ZM

B)

49

(ZW

E)%

Ag

e d

istr

ibu

tio

n o

f su

spe

cte

d

me

nin

git

is c

ase

s

AFR Sentinel Site # (Country Code)

0-5mo 6-11mo 12-23mo 24-59mo

Region # of

sites

that

met

criteria

# of

suspected

meningitis

cases that

had an LP

performed

# of LPs

performed

that have a

culture

result

recorded

mean % LPs

performed

that have a

culture

result

recorded

median %

LPs

performed

that have a

culture

result

recorded

range % LPs

performed

that have a

culture result

recorded

AFR 17 19,150 12,934 68 64 (11, 99)

AMR --- --- --- --- --- ---

EMR 11 7,323 6,294 86 93 (29, 98)

EUR --- --- --- --- --- ---

SEAR 5 6,455 5,132 80 74 (55, 85)

WPR 1 268 265 99 --- ---

Total 34 33,196 24,444 74 85 (11, 99)

34

Figure 15. Percent age distribution in EMR for 2010-2012

Figure 16. Percent age distribution in WPR for 2010-2012

Percent of suspected meningitis cases that meet the definition for probable bacterial

meningitis, 2010-2012

Probable bacterial meningitis (PBM) is defined as a suspected meningitis case with CSF

examination showing at least one of the following: 1) Turbid appearance; 2) Leukocytosis (> 100

cells/mm3); 3) Leukocytosis (10-100 cells/ mm3) AND either an elevated protein (>100 mg/dl) or

decreased glucose (< 40 mg/dl). In the absence of protein and glucose results, the first two

conditions were used for case categorization. Sentinel sites demonstrated a variable but fairly low

percentage of PBM cases among suspected meningitis cases.

0

10

20

30

40

50

60

70

80

90

100

IG (AFG) 817

(IRQ)

311

(PAK)

312

(PAK)

315

(PAK)

GG

(SDN)

KK (SDN) KO

(SDN)

AW

(YEM)

SS (YEM)ST (YEM)

% A

ge

dis

trib

uti

on

of

susp

ect

ed

me

nin

git

is c

ase

s

EMR Sentinel Site Name (Country Code)

0-5mo 6-11mo 12-23mo 24-59mo

0

10

20

30

40

50

60

70

80

90

100

8 (PNG)

% A

ge

dis

trib

uti

on

of

susp

ect

ed

me

nin

git

is c

ase

s

WPR Sentinel Site # (Country Code)

24-59mo

12-23mo

6-11mo

0-5mo

35

Figure 17. Percent of suspected meningitis cases classified as PBM in AFR

Figure 18. Percent of suspected meningitis cases classified as PBM in EMR

Figure 19. Percent of suspected meningitis cases classified as PBM in SEAR

0

10

20

30

40

50

60

70

80

90

100

6 (C

MR

)

32

(CO

D)

10

(ET

H)

14

(GH

A)

15

(GH

A)

16

(KE

N)

19

(MLI)

23

(NA

M)

28

(NE

R)

26

(NG

A)

36

(SE

N)

43

(TG

O)

45

(UG

A)

46

(UG

A)

47

(UG

A)

48

(ZM

B)

49

(ZW

E)

% P

rob

ab

le b

act

eri

al

me

nin

git

is (

PB

M)

wit

h

da

ta t

o c

lass

ify

AFR Sentinel Site # (Country Code)

Probable Bacterial Meningitis

0

10

20

30

40

50

60

70

80

90

100

IG (AFG) 817 (IRQ) 311 (PAK)312 (PAK)315 (PAK) GG (SDN) KK (SDN) KO (SDN) AW

(YEM)

SS (YEM) ST (YEM)% P

rob

ab

le b

act

eri

al

me

nin

git

is (

PB

M)

wit

h d

ata

to

cla

ssif

y

EMR Sentinel Site Name (Country Code)

Probable Bacterial Meningitis Without Probable Bacterial Meningitis

0

10

20

30

40

50

60

70

80

90

100

1 (BGD) 2 (BGD) 3 (BGD) 4 (IND) 7 (NPL)

% P

rob

ab

le b

act

eri

al

me

nin

git

is (

PB

M)

wit

h d

ata

to

cla

ssif

y

SEAR Sentinel Site # (Country Code)

36

Figure 20. Percent of suspected meningitis cases classified as PBM in WPR

Percent age distribution of probable bacterial meningitis, 2010-2012

The age distribution of PBM cases was examined by region, based on sentinel sites that reported

the age of cases during the period 2010-2012. Age distributions were observed to be relatively

more consistent compared to the age distribution of suspected bacterial meningitis cases, with

regions AFRO and WPRO showing a majority of cases in the 0-5 months category (Figure 8,

10).

Figure 21. Percent age distribution of PBM cases in AFR

0

10

20

30

40

50

60

70

80

90

100

8 (PNG)

% P

rob

ab

le b

act

eri

al

me

nin

git

is (

PB

M)

wit

h d

ata

to

cla

ssif

y

WPR Sentinel Site # (Country Code)

0

10

20

30

40

50

60

70

80

90

100

6

(CMR)

32

(COD)

10

(ETH)

14

(GHA)

15

(GHA)

16

(KEN)

19

(MLI)

23

(NAM)

28

(NER)

26

(NGA)

36

(SEN)

43

(TGO)

45

(UGA)

46

(UGA)

47

(UGA)

48

(ZMB)

49

(ZWE)

% A

ge

dis

trib

uti

on

of

pro

ba

ble

me

nin

git

is

case

s

AFR Sentinel Site # (Country Code)

0-5mo 6-11mo 12-23mo 24-59mo

37

Figure 22. Percent age distribution of PBM cases in EMR

Figure 23. Percent age distribution of PBM cases in WPR

Age distribution of laboratory-confirmed cases (Hi, Spn, Nm)

Meningitis cases that were lab-confirmed (i.e. tested positive by culture, binax, latex agglutination,

and/or PCR) were stratified by age and region for each type of vaccine-preventable disease

(VPD) organism isolated (Hi, Spn, Nm).The age distribution for each pathogen varied within

regions and between regions.

0

10

20

30

40

50

60

70

80

90

100

IG (AFG) 817 (IRQ) 311

(PAK)

312

(PAK)

315

(PAK)

GG (SDN) KK (SDN) KO (SDN) AW

(YEM)

SS (YEM) ST (YEM)

% A

ge

dis

trib

uti

on

of

pro

ba

ble

me

nin

git

is

case

s

EMR Sentinel Site Name (Country Code)

0-5mo 6-11mo 12-23mo 24-59mo

0

10

20

30

40

50

60

70

80

90

100

8 (PNG)

% A

ge

dis

trib

uti

on

of

pro

ba